Google Brings Model Context Protocol to Google Home, Opening Hardware Control to External AI Agents

Google has added Model Context Protocol (MCP) support to its Google Home ecosystem, permitting external artificial intelligence agents to access residential telemetry and operate connected smart home devices. The feature is initially expanding to Google Home Premium Advanced subscribers located within the United States, as first reported by The Verge.

The integration enables third-party AI agents to interface directly with the Google Home environment. Under the new capabilities, authorized external artificial intelligence systems can review incoming data streams generated by connected hardware throughout a user's residence.

In addition to monitoring and analyzing household information, third-party software agents connected via MCP are authorized to execute actions directly on supported hardware. This allows external platforms to trigger actions, modify settings, and manage automation sequences across the connected home network.

Google is limiting the initial availability of the MCP integration to its highest-tier subscription tier. Only customers in the U.S. who hold an active Google Home Premium Advanced plan are currently eligible to configure and utilize external AI agent access.

The adoption of the Model Context Protocol reflects a broader effort within consumer technology to standardize how generative and autonomous AI tools communicate with Internet of Things (IoT) hardware. By opening its smart home infrastructure to standardized protocol calls, Google is facilitating broader interoperability between rival software agents and its underlying hardware ecosystem.

The release details were originally reported by Jennifer Pattison Tuohy of The Verge and documented in industry updates on September 16, 2026. Google has not yet detailed expanded availability timelines for standard tier subscribers or international regions.